tilearn API reference

The tilearn package provides pure-Python helpers for single-machine scheduling workflows, including classic sequencing heuristics, factor calculations, and multi-CSV orchestration.

Data model contract

Most APIs operate on a job table represented as:

[
    [name, p, r, d, w, ...],
    ...
]

• name: job identifier (str)
• p: processing time (float)
• r: release time (int)
• d: due date (int)
• w: weight (float)

Many functions append computed columns in-place (for example, w/p, completion time, or lateness).

In-place mutation

Several functions mutate rows directly. If you need to preserve your original input, create a deep copy before calling scheduling helpers.

Module map

• tilearn.bs.basis: core matrix helpers, factors, and schedule metrics.
• tilearn.wspt and tilearn.edd: sequencing rules for weighted completion and maximum lateness workflows.
• tilearn.data: CSV backup/read/update/cleanup helpers.
• tilearn.joblist.plat and tilearn.joblist.run: multi-list orchestration APIs and end-to-end optimal_list pipeline.

Quick start

import tilearn as tl

jobs = [
    ["J1", 2.0, 0, 8, 4.0],
    ["J2", 1.0, 0, 5, 1.0],
]

ranked = tl.wspt(jobs)

For guided tutorials and theory-first explanations, see the User Guide.